Generally, PET can be produced in two steps by one of two ways, called the DMT and the TPA processes, or the transesterfication and direct esterification routes, respectively. Since its commercialization in the 1950’s by DuPont, PET was predominantly synthesized out of DMT due to the fact that this monomer, compared to TPA, was relatively easily purified by distillation. Yet, since the 1960’s, purified TPA became available by the development of new technologies and gained a lot of importance as a monomer in polyester synthesis (Matsuzawa, K., 1976). The applications for PET include fibres and filaments, films, and bottlegrade chips. Modem plants are based on the TPA process and further, they incorporate direct product formation (fibres and filaments, films) by extruding the melt from the final polycondensation reactor.
First we shall contrast the DMT and the TPA processes. The main difference is the starting material. The older process used dimethyl terephthalate (DMT) and ethylene glycol (EG) as starting materials. This was because of the non availability of terephthalic acid of sufficient purity in the early years of polyester production. The newer industrial method uses purified terephthalic acid (TPA) instead of DMT and is called the TPA process. TPA was introduced on a commercial scale for the first time in the world by Amoco Corporation, USA in 1963. Since then, several plants for producing TPA have been set up all over the world. TPA has significant advantages over DMT and hence world over it has become the preferred raw material, in comparison with DMT, for polyester production.
The major advantages of TPA over DMT are as follows:
* Weight of TPA required per ton of fibre is about 15% less than that of DMT. * Lower TPA: glycol ratios are possible as compared to DMT: glycol ratio. * Methanol as a by-product is not produced when TPA is used as in DMT use. Hence, no methanol recovery plant is needed. * No-catalyst is required for TPA esterification whereas it is required in DMT transesterification.
Reaction byproduct and catalyst
In the DMT process, in the first step, DMT is trans-esterified with ethylene glycol (EG) to produce an intermediate called diethylene glycol terephthalate (DGT) plus a small amount of low oligomers. The reaction byproduct is methanol and this is distilled off. The DGT is alternatively called bis hydroxy ethyl terephthalate or BHET in the literature. Manganese (II) acetate or zinc (II) acetate is typically used for this transesterification step, these being the best catalysts for this reaction. In the second step, the DGT is heated to about 280°C under high vacuum to carry out melt-phase polycondensation. The principal volatile eliminated is EG. For the second step in the DMT route, the catalyst from the first step (zinc or manganese) is sequestered or deactivated with phosphoric acid (U.S. patent 5898059), and another catalyst for polycondensation, most commonly antimony triacetate or antimony trioxide is added. This is because zinc and manganese are considered poor polycondensation catalysts.
The literature indicates that the reactivity of metals for the polycondensation reaction (second step) follows the trend Ti > Sn > Sb > Ge > Mn > Zn (T.H. Shah, J.I. Bhatty, G.A. Gamlen and D. Dollimore, Polymer, 25, 1333 (1984)). Moreover, Shah et al. indicate that for the first step, namely the transesterification of DMT with EG, the catalytic activity trend follows the reverse order, with zinc being amongst the most active. For the polycondensation reaction, Sb compounds are commercially established (compared with Sn and Ti) because the resulting polymer has the most favourable balance of properties. Note, in a usual operation, it is possible to go from step 1 to step 2 without isolating the DGT. However, if desired, the DGT and oligomers formed in step 1 can be isolated and used later for melt polycondensation (step 2).